feat(tree): Add distance matrix to newick conversion method.

docs/source/index.rst

@@ -51,6 +51,7 @@ This has been written with the intention of personal use, but feel free to use/c
 
    util/io
    util/accession
+   util/tree
 
 
 .. toctree::

docs/source/util/tree.rst

@@ -0,0 +1,5 @@
+****
+Tree
+****
+
+.. autofunction:: magna.util.tree.dm_to_newick

magna/util/tree.py

@@ -0,0 +1,31 @@
+from typing import Collection
+
+import numpy as np
+from scipy.cluster import hierarchy
+
+
+def _get_newick(node, parent_dist, leaf_names, newick='') -> str:
+    if node.is_leaf():
+        return "%s:%.2f%s" % (leaf_names[node.id], parent_dist - node.dist, newick)
+    else:
+        if len(newick) > 0:
+            newick = "):%.2f%s" % (parent_dist - node.dist, newick)
+        else:
+            newick = ");"
+        newick = _get_newick(node.get_left(), node.dist, leaf_names, newick=newick)
+        newick = _get_newick(node.get_right(), node.dist, leaf_names, newick=",%s" % newick)
+        newick = "(%s" % newick
+        return newick
+
+
+def dm_to_newick(arr: np.ndarray, labels: Collection[str]) -> str:
+    """Convert a pairwise distance matrix into Newick format.
+
+    Args:
+        arr: The symmetrical pairwise distance matrix.
+        labels: The labels of the leaf nodes.
+    """
+    Z = hierarchy.linkage(arr)
+    tree = hierarchy.to_tree(Z)
+    newick = _get_newick(tree, tree.dist, labels)
+    return newick

setup.py

@@ -44,6 +44,6 @@ def readme():
       packages=find_packages(),
       include_package_data=True,
       install_requires=['tqdm', 'pandas>=1.1.0', 'pyarrow', 'numpy',
-                        'dendropy', 'biopython'],
+                        'dendropy', 'biopython', 'scipy'],
       python_requires='>=3.6',
       )